914 THE AMERICAN NATURALIST.: {NOL XXXII. 
ivy was from Kenilworth, Abbotsford, and Dryburgh, and was 
packed in earth at Queenstown. The garden where the snails 
were first noticed is about 200 yards from my garden. They 
spread at first with considerable rapidity, but since the first few 
years their limits have widened only slowly. As far as I can 
learn, by examination and inquiry, they are confined to a territory 
not over one and one-fourth miles long and one-half mile wide. 
My aim in collecting the three series, which I hope at some 
future time to supplement, was to get light on the following 
questions : (2) Does the tendency to variation proceed along cer- 
tain definite lines, and if so, what? (4) Does this tendency vary 
in different localities in the colony? (e) Will a very consider- 
able destruction of individuals modify materially this tendency? 
A study of these three series seems to give an affirmative 
answer to “a” and “6,” and a negative answer to s.. 
The succeeding tables, which summarize the results of these 
series, have reference solely to the danding of the shell; of 
series A, 87%, and of series B, 83% are var. /ébellula (with 
yellow ground). In series “ C” every shell but one is de//u/a. 
The formule used in designating the varieties are those 
which have of late been generally adopted, and for a knowledge 
of which I am indebted to 
Professor Cockerell. The 
typical shell of the Lex- 
ington colony, or at least 
Fines the most common variety 
Shell of Helix nemoralis. The bands are numbered r to 5. (libellula), has a yellow 
The right-hand figure shows bands 4 and 5 only. groun 4 oe à ree ban ds, 
three above the periphery and two (usually broader bands) below. 
These are numbered from the top of the body whorl (near the 
mouth) downward 1 2 3 4 5, as shown in the accompanying cut. 
A band which is incomplete or appears only in traces, t.2., rudi- 
mentary, is designated in the formule by a small number, as 
12345. A band which is wanting is designated O, as in 10345. 
Obviously there will be every intergrade between the formule, 
as between 12345 and 12345. In some European shells a band 
is interrupted at intervals, and for this a colon (:) is used, but 
these have not been noticed in Lexington. When two bands 
